Writing and Literature 59 ink from Beth Shemesh, the characters upon which are certainly archaic but not pictorial. Of greater importance is the series of inscriptions discovered by the foully murdered excavator Starkey at Tell el-Duweir (Lachish). Of these the oldest is a dagger bearing four letters placed vertically (see Plate I, fig. 2), the second of them a clear head like that common at Serabit, and the third possibly but not certainly a snake. The dagger is dated by Starkey 'perhaps before, but not later than 1600 B.C.' Next in order of development is a ewer bearing ten or eleven letters clearly less pictorial, which Starkey placed without hesitation at the beginning of the thirteenth century B.C. (Plate I, fig. 3). Several of the signs correspond pretty obviously to Serabit characters on the one hand, and to Phoeni- cian letters on the other. The zigzag for m, the cross for t, and the snake for n are confidently interpreted by many as the common Hebrew word mattdn 'gift of . . .', and the probability of this interpretation is enhanced by the presence, after these letters, of three dots, obviously a word-divider as in certain early Greek inscriptions. Further along are a Idmsdb with the crook at the bottom as in Phoenician and Latin, and an dlepb that no longer much resembles an ox-head, but has affinities with the dleph of the Moabite Stone on the one hand, and with a Greek alpha on the other. A third text from Tell el-Duweir (PL I, fig. 4)1 has a combination of five letters followed by a word-divider like that on the inscription of Ahiram (see below), and Hebraists and Serabitists alike have joined hands in reading these letters as besbelosbetb Tor the third time'. Here we are nearer to Phoenician both in graphic development and in date; Starkey assigns the bowl to the third quarter of the thirteenth century, Yeivin to the fourth. If the first character on the bowl (^ be really beih, then the divergency 1 Doubts have been felt as to which way up this inscription should be placed. I have followed the opinion of J. Obermann, who places it in such a way that it reads from left to right.